Edge Computing for Real-time Vision Applications: Investigating edge computing techniques for real-time vision applications, including object detection and surveillance systems

Abstract

Edge computing has emerged as a promising paradigm for enabling real-time processing and analysis of data generated by devices at the network edge. In the context of vision applications, such as object detection and surveillance systems, the need for low latency and efficient utilization of network resources makes edge computing a compelling solution. This paper provides an overview of edge computing techniques tailored for real-time vision applications. We discuss the challenges and opportunities in implementing edge computing for vision tasks and review existing approaches and frameworks. Additionally, we present a comparative analysis of these techniques based on their performance, scalability, and resource efficiency. Our findings suggest that edge computing can significantly enhance the performance of real-time vision applications by offloading computational tasks to edge devices, reducing latency, and improving scalability. We conclude with future research directions and open challenges in the field.